Inter-rater reliability of fish vitality assessments: does the use of categorical or continuous scoring scales affect the Reflex Action Mortality Predictor (RAMP)? "Pieter Meeremans"

Description

To incentivise fishermen to fish more selectively, the revised Common Fisheries Policy (CFP) includes an obligation to land all catches of quota-regulated species. From January 1st, 2016, it applies to some demersal fisheries and their target species and by 2019 to catches of all quota-regulated species. To stay viable, those fisheries that struggle to improve selectivity, may seek exemptions from this obligation. One such exemption requires scientific evidence to demonstrate "high survival" of discarded fish. This evidence can be provided by using the Reflex Action Mortality Predictor (RAMP) approach, which relates vitality status indicators such as reflex responses and/or injuries to mortality probabilities. However, when several raters with different backgrounds, levels of experience or training become involved in scoring reflexes and/or injuries, there is potential bias from subjectivity in measurement. To improve credibility of results and to maximise accuracy, it is important to assess inter-rater reliability and to evaluate whether any variability affects significance (if any) of a vitality-mortality relationship. Therefore, reflex and injury assessments of 304 beam-trawled plaice were performed by 3 different raters during four trips on-board the RV Simon Stevin using both a 3-point categorical scale and a tagged visual analogue scale. In contrast to categorical injury scores and continuous reflex scores, categorical reflex scores were not significantly different among raters. Even though categorical reflex indices, reflex & injury indices and continuous indices were significantly different among raters, high inter-rater agreement was achieved for the indices on both scales. Best fits for the relation between indices and mortality were observed for both continuous reflex indices and categorical reflex & injury indices. Sufficient training, experience and good discriminative power seemed to be important requisites for a rater to produce a good fit between vitality status and delayed mortality. This study will make it possible to fine-tune the rating criteria and rating system in the future. Although the results of this study seem to favour the use of a continuous scale for fish vitality assessments, a final choice will depend on further research concerning the effect of fish size and the applicability under different weather conditions.

Period	2015 → 2016
Degree of Recognition	International